Following back to back user events and electing to drive between San Francisco, Simi Valley, and Las Vegas so as to not have to face long check-in and security lines, I took the opportunity to take some time out racing cars again with the new Connect Vice President, Margo Holen.
And I have included a picture of her just prior to a group "download" session where the lead instructor critiques us after being on the track. I have to say, Margo seems very much in her element - and, as I once heard in a movie, I complemented her on her nice driving shoes! Color-coordinated and all.
This was our second outing as we worked our way up from the lower novice rankings, and this one proved to be a real challenge. Prior to arriving at the ButtonWillow track, we had gone to the web and printed off a very detailed track report provided by the Porsche club. It described the lines through every corner and identified landmarks that would help with turning into a corner. It even gave advice about which corners to just forget about as they were less important than the corners that followed. In one four-corner sequence, you had to drive off the racing line in order to be in the right place for the last two corners and then have an opportunity to get the best from your car in the short straight that followed. Giving up on a corner, any corner, just didn't seem right.
However, all our planning was for nothing as the first news from our instructor Fulton was "today, we will be driving the circuit counter-clockwise!" As I drove out onto the course for the first time, and with no real sense of where it went, I was soon to discover that this was the most difficult track of all to drive, and technically required drivers to adapt to many new combinations of corners and to handle a number of undulations that unsettle the car. Adding to each drivers stress levels, the weekend was particularly hot with temperatures climbing as high as 110 F!
But the cars were OK. They ran a little hotter but they were never the problem - the cars still needed their drivers input and it was up to the drivers to adapt to the track, and to corners that were unfamiliar, to the extremely hot conditions and to the dirt that was thrown up as cars put two wheels off the track. By midday Sunday, the attrition rate had climbed to where it looked as though only half the drivers remained. Those drivers uncomfortable with the conditions and not prepared to continue, elected to leave the weekend's program.
The weekend came as I was preparing to fly to London. All part of a schedule that I knew would take me away from home for three weeks. But the weekend break was a time to clear the mind, and to focus on something else entirely, but when I returned to Simi Valley I was as mentally exhausted as I was physically and some five pounds lighter.
Monday morning had me wrapping up loose ends, and attending to email correspondence that I had left unanswered during HP's Technical Forum and Expo (HPTF&E). Late in the morning I received a phone call and, as I answered, a synthetic voice began "this is a United 'cancelled flight' status update - your flight between Los Angeles and Denver has been cancelled!" I needed this flight, as my flight to London was a new service out of Denver and I had negotiated a good promotional fare. Compounding the issue, all other flights to Denver were full and, "would I consider flying another day", was the suggestion provided by the reservation agent I called.
Thinking quickly and once again, adapting to the new conditions, I began to negotiate with the reservations agent. Can I fly direct to London out of Los Angeles? No - that flight is boarding. Can I fly to San Francisco? Again, no seats on the connecting flights. How about Burbank to Denver? Yes, there's one seat left but you need to be at the airport in one hours time - and with normal driving conditions, the trip to Burbank from Simi Valley is about 40 minutes. With few other options, I took the seat and threw everything I needed into a bag and made it to the airport on time.
I always take time to plan before leaving - I go to mapquest and print off maps of where my hotels are located and where all meetings were to be held. I always know where I am headed. However, these days and with the turbulent times the airlines are having, nothing can be taken for granted. I don't have it all figured out, all of the time, but I am really having to become very flexible and learn to be a whole lot more patient, when it comes to travelling these days.
Arriving in London, I had thought I had put all the changes behind me but no sooner had I taken a seat on the Heathrow Express to Paddington than the conductor came down the aisle opening windows - apologising for an air conditioning failure that the operator hoped to correct once they returned to Paddington. London was going through a period of warm weather and the train was a hot-house but with the windows down, it remained a hot-house only a lot more noisy! I switched to London's famous Underground - only to find the timetables disrupted as "there is a person on the line" according to the announcement over the PA system. Finally, I arrived at St Pauls station only to find the escalators inoperable with a sign saying "sorry for any inconvenience, but the station is undergoing a major infrastructure upgrade!"
With the HPTF&E sessions still in the back of my mind, I recalled a number of discussions I had participated in when it came to planning server and operating system upgrades. This was particularly relevant as HP had announced that NonStop would be supported by Blades and the Integrity NonStop BladeSystem would be the first offering in this new product family. Much anticipated, and openly depicted in product roadmaps for some time, it was still an exciting development for all event participants.
While some event participants have expressed concern over the shortened product lifecycles and the availability of ever-more-powerful products, for most of us having access to these new servers would facilitate even more applications deployment on NonStop, with even greater savings. But integrating a new server into an existing fabric of servers, and to do so with no downtime for mission critical applications, has now become an extremely important consideration for the data center manager keeping in mind the Service-Level Agreement (SLA) he has to meet. For them, maintaining fall-back positions every step of the way has reached new levels of criticality!
But even here, these data center managers have to remain flexible, and be able to adapt to changes that happen at the last minute and often without warning. No matter how many migrations are planned, there's always something unexpected occurring that forces adjustments along the way.
In today's IT world the design of the data center, and the way everything is connected and tightly integrated, has to take into consideration that change will be ongoing. There's just so many components these days scattered across not just two data centers, but increasingly three and often as many as five or six data centers, as enterprises operate globally. During HPTF&E, HP's CIO Randy Mott provided a video clip of how the HP data centers have been set up under his direction. Six separate facilities spread across three cities. Each data center built with deep false flooring for power and cooling, overhead gantries carrying the wiring, networking facilities with switches and routers isolated from the rest of the equipment, all well thought out and deliberately constructed to accommodate an active life full of continuous and ongoing upgrade programs.
And the transition to new servers small or large, is not a one-way street even for sites that are as big as HPs! At every step along the way, the ability to switch the application back to the original server, with no loss of data or interruption of the services, should always be a part of the plan. Having the tools to keep data bases, for instance, current and synchronised on both the old and the new server, just has to be taken in consideration as you never know when management elects to reverse the direction of the transaction traffic and falls back to the older server.
Smart data center managers are rapidly moving away from the old Disaster Recovery (DR) site and fully integrating within their operational environment so that there's never a concern about a site's "currency". If it is an active participant in the overall operational environment, then there is less concern about directing the transaction traffic at it as the applications are all "live" with minimal exposure to an older code base or data base release.
Driving back to Simi Valley after a weekend at the track, I heard a new song from Alanis Morissette, where she sings:
"And what it all boils down to, is that no one's figured it out just yet ... 'cause i've got one hand in my pocket - and other is hailing a taxi cab"
Turning up at a race track and finding the direction had been reversed! Getting ready to head overseas to find I had to rush to a different airport! Rushing into London with the infrastructure failing and in the middle of upgrades! Rolling in a new server only to find that changes were made to the application code after copies were made, or tables in the data base changed, or any one of the networked target servers being integrated going down unexpectedly all forces us to adapt.
What it all boils down to is that for sure I haven't got it figured out - and if United keeps cancelling flights and laying off pilots, and London's infrastructure makes rail travel an unreliable and miserable experience then perhaps, as Alanis sings, I will hailing a taxi to my next meeting ... wherever!